Broker comparisons are conducted in pips and the results are experienced in currency. The translation between them is annual volume, and because most traders do not know their own volume with any precision, the comparison never becomes concrete. Computing the number once changes several decisions at the same time.

The five components

  • Spread paid, across every entry and exit.
  • Commission, where the account charges it separately.
  • Overnight financing on positions held past rollover.
  • Currency conversion on settlements outside the account currency.
  • Account charges: inactivity, withdrawals, platform or data fees.

Start with volume, not with rates

Export a year of closed positions from the platform. Sum the volume in standard lots — the platform reports it, or it is the sum of position sizes. This single figure is what converts every rate into money, and it is almost always different from what the trader expected.

Traders who run this step commonly find their annual volume is two to three times what they assumed, because small positions accumulate invisibly and because round trips count twice in cost terms. A method that felt like twenty trades a month is often forty round trips and eighty spread payments.

Spread and commission

Multiply the volume by the average spread you actually paid on the instruments you traded, converted to money at the pip value. Do not use the advertised spread — take it from your own fills, or from a sample of spread readings at the hours you trade.

Add commission directly: it is stated per lot and needs no conversion. If the account is all-in, this term is zero and the spread term is correspondingly larger.

This subtotal is what every broker comparison discusses, and for a short-term trader it is most of the answer. For anyone holding positions overnight it is not.

Financing

Sum the swap and financing lines from the statement. This is the term traders most often omit, and on a position-holding strategy it frequently exceeds spread and commission combined.

Financing on CFDs is charged on the full notional rather than the margin, so an index or commodity position of modest margin can accumulate a charge out of proportion to the account. Weekends count three days, and a position held through a long holiday accumulates across all of it.

The statement has the number; you do not need to model it

Every broker reports swap and financing as separate lines. Summing a year of them takes minutes in a spreadsheet and produces a figure no amount of reading about swap rates will give you.

Currency conversion

This is the term that does not appear on the statement at all. Every trade settled in a currency other than the account's is converted at a marked-up rate, embedded in the rate rather than shown as a fee.

Estimate it: take the total absolute value of results settled in other currencies — the sum of wins and losses without netting — and multiply by the markup from the fee schedule. At 0.5% on twenty thousand units of gross settlement, that is a hundred units a year, invisible everywhere.

Add deposit and withdrawal conversions separately, since those are usually applied by your bank or card issuer rather than the broker and appear on a different statement entirely.

Account charges

Inactivity fees, withdrawal charges, platform subscriptions, market data fees where they apply. Individually small; collectively worth counting, particularly for a trader holding accounts at several brokers where the dormant ones are quietly being drawn down.

What to do with the total

Turning the number into decisions

  1. Express total annual cost as a percentage of average account equity.
  2. Compare it against your annual return — if cost is a large fraction of gross profit, the strategy is fragile.
  3. Identify which component dominates: it is rarely the one you were optimising.
  4. Recompute the total under an alternative account type or broker using the same volume.
  5. Recompute under a reduced trade frequency, which is usually the largest available saving.
  6. Repeat the exercise annually rather than once.

The third step is where the exercise earns its keep. A trader spending months comparing spreads while paying three times as much in financing has been optimising the wrong term, and the total is what reveals it.

The uncomfortable finding

For many retail traders the largest single lever is not the broker at all. It is trade frequency: halving the number of round trips halves the spread and commission terms directly, and it usually improves the quality of the remaining trades because the marginal ones were the weakest.

That is not an argument for a worse broker. It is an argument for computing the total before assuming the broker is where the money is going.

Common questions

What is a normal annual trading cost?

There is no benchmark, because it depends entirely on frequency and holding period. The useful comparison is against your own gross profit: cost as a large fraction of it means the method has little margin for error.

Which cost do traders most often overlook?

Currency conversion, because it never appears as a line item — it is embedded in the rate applied to each settlement. Financing is second, and it dominates for anyone holding positions overnight.

Should I switch brokers to save on cost?

Only after computing the saving at your actual volume. A difference of 0.3 pips is worth $3 per standard lot; whether that justifies moving depends on how many lots you trade, and on how the two brokers compare on withdrawal reliability.

How often should I recompute this?

Annually, and after any material change in method, instrument or account type. The components shift more than traders expect when a strategy's holding period changes.

Related reading